A simple 16-bit architecture, with an emulator
N/A
You can build the packages individually by invoking make in their directories.
The following are required to build all packages:
make
gcc
gperf
A, B, C, D
IP, SP, FLAGS
INSTRUCTION | DESCRIPTION | OPCODE |
--------------------------------|------------------------|---------|
CMP (R1) (R2) -> FLAGS | (R1) - (R2) -> FLAGS | 0x0 |
ADD (R1) (R2) -> (R1) | (R1) + (R2) -> (R1) | 0x1 |
SUB (R1) (R2) -> (R1) | (R1) - (R2) -> (R1) | 0x2 |
JNZ (IMM/REG) | (IMM/REG) -> (IP) | 0x3 |
PUSH (REG) | (REG) -> (--SP) | 0x4 |
POP (REG) | (SP++) -> (REG) | 0x5 |
ST (IMM/REGPTR) (REG) | (REG) -> (MEM) | 0x6 |
LD (REG) (IMM/REGPTR) | (MEM) -> (REG) | 0x7 |
OR (R1) (R2) -> (R1) | (R1) | (R2) -> (R1) | 0x8 |
AND (R1) (R2) -> (R1) | (R1) & (R2) -> (R1) | 0x9 |
XOR (R1) (R2) -> (R1) | (R1) ^ (R2) -> (R1) | 0xA |
LSH (R1) (R2) -> (R1) | (R1) << (R2) -> (R1) | 0xB |
RSH (R1) (R2) -> (R1) | (R1) >> (R2) -> (R1) | 0xC |
CLI | FLAGS &= (~I) | 0xD |
STI | FLAGS |= (I) | 0xE |
INT (IMM) | | 0xF |
0x1: Zero
0x2: Negative
0x4: Overflow
0x8: Interrupt Enable
16-bit addressable memory
All instructions are of 16-bit length, stored in memory in little-endian format
After reading from memory, we decode the instruction into the following format:
AAAA RRRR BBCC MRRR
A: Instruction
B: R1
C: R2
M: Mode (1 for immediate)
R: Reserved
In the case of an immediate value, the immediate value is stored in the next two bytes of memory, again in little-endian format. Due to the way the opcode is decoded, this format allows for duplicate opcodes.